Luciferases are a class of proteins expressed in a wide variety of terrestrial and marine animal and plant species that enable communication of signals via bioluminescence. One of the most familiar such systems is known from the American firefly, Photinus pyralis. The enzymatic activity of luciferase catalyzes the reaction of ATP, oxygen and the cofactor, luciferin, releasing energy as a visible photon together with the byproducts, AMP, pyrophosphate and oxidized luciferin. Chemiluminescence is one of the most sensitive ways to probe molecular mechanisms because, (1) the conversion of the biochemical energy of ATP into light by this enzyme has a quantum yield of approximately 90%, and (2) single photon counting can be performed using advanced low-light-level detection systems, enabling the recording and study of events from single molecule reactions. As noted above, there are unexplained complexities in the behavior of luciferase, and in particular, we have noted certain paradoxical levels of light production and non-linearity occur when the enzyme is exposed to its known substrate, ATP, in the presence of other nucleotides, such as ADP. Indeed, it has been reasoned that much of this mechanistic uncertainty could be explained by a certain level of luciferase contamination with adenylate kinase (a.k.a., "myokinase"), which is abundant in preparations from which luciferase is purified (e.g., contaminated by the flight muscles in fireflies, as well as from the cytoplasm of bacteria used in recombinant protein generation, etc.). Adenylate kinase produces ATP through the dismutation of ADP, which would produce an unforseen and uncontrolled contaminating light signal in the luciferase reaction independent of the ATP originally present and thus confounding the accuracy and specificity of the measurement. We sought to examine the catalytic activity and chemiluminescence mechanisms of luciferase in the presence of ADP. We found that under certain circumstances where ADP is in quantitative excess over ATP, there can be a significant excess light output vs that expected quantitatively from pure ATP standards. To rule out that adenylate kinase contamination was responsible for this apparent artifact, we purified several luciferases (both from firefly extracts and recombinant material) using preparative column chromatography designed to exclude <30 kD species (adenylate kinase mw approx. 20 kD). After confirming the quantitative depletion of species <30 kD, we found that the ADP-related artifact on luciferase light output was still present, indicating that adenylate kinase contamination was not the cause. It has been reported elsewhere that inorganic pyrophosphate (PPi) may increase the output of luciferase bioluminescence. To rule out that the possible activating effect of ADP in our bioluminescence assay was an artifact caused by the presence of PPi contamination in the preparations used, we treated the ADP with inorganic pyrophospatase (PPiase) that converts one molecule of pyrophosphate to two phosphate ions. PPiase treatment although changing the kinetics of the light production had no effect on the total light output. Furthermore, we demonstrated that addition of external PPi to ADP did not activate but instead significantly inhibited luciferase light production and this inhibition was reversed by treatment with PPiase. Thus, we ruled out an artifact produced by PPi contamination. We then hypothesized that luciferase itself might be able to obtain sufficient energy from ADP alone to produce the light-output artifact. In order to test this, we developed an in-gel chemiluminescence assay run on a clear native gel platform. The luciferase was first separated from other potential enzyme contaminants using a clear native gel. This gel was subsequently incubated with purified ADP and luciferin, and low light level imaging showed that 2 discrete bands yielded significant light output. Subsequent addition of an excess of ATP to this mixture showed that these exact same bands that produced light with ADP alone, were also producing substantially increased light with the (excess) ATP. At the end of the imaging procedure, immunoblotting of this gel confirmed that the bands producing light were positively stained for luciferase (and negative for adenylate kinase), and proteomic analysis of these bands (from gel lanes not used for immunoblotting) identified that the only protein present was luciferase without any other protein contamination. We conclude that luciferase itself is capable of using the energy of ADP to produce light output independently of the presence of preformed ATP, possibly by the catalytic conversion of ADP to ATP and the latter being utilized as the ultimate substrate.